A simple "how many times per day" question may be better than detailed daily logs for measuring general cannabis exposure | Research

The takeaway

Comparing two cannabis measurement methods in 1,090 regular users, a brief quantity-frequency scale outperformed detailed daily logs for predicting the THC metabolite marker of general exposure, while daily logs better predicted acute THC levels.

Cannabis researchers designing studies, clinicians assessing patient use, and methodologists interested in substance use measurement.

Simple frequency measure outperformed daily logs for general exposure (R² 0.30 vs 0.27)

What the researchers found

The Cannabis Quantity and Frequency Scale (CQFS) total times/day metric was a better predictor of the long-term THC metabolite (THC-COOH, R-squared 0.30 vs 0.27) while the Timeline Follow Back (TLFB) days/month metric was better at predicting acute blood THC levels (R-squared 0.24 vs 0.21).

Why it matters

Cannabis research is hampered by inconsistent measurement of use. This study provides empirical guidance on which measurement tool to choose depending on the research question, potentially standardizing an area that has long lacked consensus.

The numbers in context

1,090 participants, mean age 32.89. Average use: 16 days/past month, 4 times/day. 78.35% White, 51.56% female. CQFS model R-squared for THC-COOH: 0.30 vs TLFB 0.27. TLFB model R-squared for THC: 0.24 vs CQFS 0.21. CQFS total times/day predicted THC-COOH (B=5.85, p=.03).

How the study worked

Observational study pooling data from five larger studies in the Boulder/Denver area. 1,090 regular cannabis users completed the CQFS (typical quantity/frequency) and TLFB (past-month daily use). Blood biomarkers (THC after acute use and baseline THC-COOH) were collected for comparison.

What this study cannot tell us

Colorado-based sample of regular users may not generalize to occasional users or other regions. Predominantly White sample. Biomarker collection at a single time point. The five pooled studies may have had varying protocols.

How to read the evidence

Moderate: large sample with biomarker validation, but single-region convenience sample pooled from multiple studies with potentially varying protocols.

When this study was published

Published 2026.

The bigger picture

As cannabis research scales up, having validated, efficient measurement tools becomes critical. This study suggests researchers can choose simpler instruments for some applications without sacrificing accuracy, potentially improving participation and data quality.

Questions still open

Should cannabis research standardize on one measurement approach? How do these measurement tools perform in populations with different use patterns? Does the sex difference in CQFS-THC prediction reflect biological or behavioral factors?

Common questions

What is the best way to measure cannabis use in research?

This study suggests a simple "total times per day" question better captures general cannabis exposure, while detailed daily logs are better for measuring acute use patterns.

Does sex affect cannabis measurement accuracy?

The brief frequency measure predicted acute THC levels in males but not females, suggesting sex-specific patterns in how use frequency relates to blood THC levels.

Read the original research

Cannabis use measurement: Identifying the optimal metric for broad research applications.

Addiction (Abingdon, England), 121(2), 331-339

View on PubMed ↗Publisher’s paper ↗

Citation

Skrzynski, Carillon J; Mueller, Raeghan L; Bidwell, L Cinnamon; Bryan, Angela D; Hutchison, Kent E. (2026). Cannabis use measurement: Identifying the optimal metric for broad research applications.. Addiction (Abingdon, England), 121(2), 331-339. https://doi.org/10.1111/add.70205